FREE Registration is required
Overview:
This paper addresses the problem of extracting key pieces of information from voicemail messages, such as the identity and phone number of the caller. This task differs from the named entity task in that the information this paper is interested in is a subset of the named entities in the message, and consequently, the need to pick the correct subset makes the problem more difficult. Also, the caller's identity may include information that is not typically associated with a named entity. This paper presents three information extraction methods, one based on hand-crafted rules, one based on maximum entropy tagging, and one based on probabilistic transducer induction. It evaluate their performance on both manually transcribed messages and on the output of a speech recognition system.
(Is this item miscategorized? Does it need more tags? Let us know.)
| Format: | Size: | 93 KB | |
| Date: | Jan 2008 | ||
| Pages: | 8 |
People who downloaded this item also downloaded
Top results from Voice Mail
White Papers, Webcasts, and Resources
- The Business Benefits of Secure Remote Access SonicWALLExamine today's top trends in telecommuting, the role of Virtual Private Networks, and how to evaluate a secure remote access vendor.
- Staffing Service Coordinates Sales Activities, Utilizes Business Intelligence with CRM Solution MicrosoftFollow one company's search for a CRM solution to help them more effectively coordinate across numerous accounts with multiple sales opportunities.
- Forrester Best Practices: Improve IT Service Delivery and Support Citrix OnlineFind out what industry watchdog Forrester has identified as five best practices for improving IT service delivery and support. (Citrix Online)
Premier Vendor Content Whitepapers, webcasts & resources from our Power Center Sponsors
- News, Insights, Guidance
Visit CBSMoneyWatch.com Today -
MoneyWatch.com is the premier destination for smart, practical personal finance advice. Watch the latest Human Capital videos to make the most of your biggest asset - your earning power
- Learn more >>
Featured Training Courses
Meet Doc
-
Here to help you with your Document Management Needs
- Doc is an enigma. Born to a Russian ballerina and a German electrical engineer, he grew up in various locations in the United States. He’s seen the insides of more brands, versions, and generations of printer and printer-related hardware than almost anyone.
- To learn more about this mysterious figure check out his blog on ZDNet and his Workspace on TechRepublic. You’ll be glad you did.
-
Produced by
ZDNet and






